1.) Raw data sourced from Nick Bloom's website
http://www.stanford.edu/~nbloom/index_files/Page315.htm
"Identifying Technology Spillovers and Product Market Rivalry"

subsidiaries_revenue_bloom.dta
subsidiaries_patents_byyear.dta
r and d data_bloom.dta

2.) We do not include intermediate datasets generated by the R and .do file below.
The final dataset is:
spillover_final2.dta

3.) R scripts should be run in the following order:
1 - Jaffe distances.R
2 - Mahalanobis distances.R
3 - Organize outputs.do - All results are generated in the R scripts. This STATA .do file combines a large number of datasets to generate spillover_final2.dta.
4 - Results with varying tech.R
5 - Measurement error bsic.R

